Azure eBook series

Cloud
migration
essentials
A guide to migrating servers
and virtual machines

1
Information in this document, including URL and
other internet website references, is subject to change
without notice. Unless otherwise noted, the companies,
organisations, products, domain names, email
addresses, logos, people, places and events depicted in
examples herein are fictitious. No association with any
real company, organisation, product, domain name,
email address, logo, person, place or event is intended
or should be inferred. Complying with all applicable
copyright laws is the responsibility of the user. Without
limiting the rights under copyright, no part of this
document may be reproduced, stored in or introduced
into a retrieval system, or transmitted in any form or
by any means (electronic, mechanical, photocopying,
recording or otherwise) or for any purpose, without the
express written permission of Microsoft Corporation.
© 2018 Microsoft Corporation. All rights reserved.
Microsoft, Azure, Active Directory, Dynamics,
ExpressRoute, Hyper-V, SQL Server, SharePoint, Windows
and Windows Server are either registered trademarks
or trademarks of the Microsoft group of companies.
All other trademarks are property of their respective
owners.
GARTNER is a registered trademark and service mark
of Gartner, Inc. and/or its affiliates in the US and
internationally, and is used herein with permission.
All rights reserved.

Microsoft may have patents, patent applications,

trademarks, copyrights or other intellectual property
rights covering subject matter in this document. Except as
expressly provided in any written licence agreement from
Microsoft, the furnishing of this document does not give
you any licence to these patents, trademarks, copyrights
or other intellectual property.

2
Contents

Chapter 1: Introduction 4 Chapter 4: Migrate 17

Why migrate now? 5 Real-time replication 18

Testing 18
Chapter 2: How do you approach Taking further advantage of your
cloud migration? 6 cloud provider to drive application innovation 19

Before you migrate 11 Tools for migration 20

Identity 11

Storage 12 Chapter 5: Optimise 21

Networking 12 Secure cloud resources 21

Connectivity 12 Protecting data 22
Readiness 13 Monitoring cloud health 22
Beginning your migration 13 Continual cost efficiency and optimisation 24

Tools for optimisation 24

Chapter 3: Assess 14
Assess in four steps 15

Assess on-premises applications and servers 15 Chapter 6: Summary 25

Identification of application and

server dependencies 15

Configuration analysis 15

Cost planning 16

Tools for assessment 16

3
 01
Introduction
In the past few years cloud adoption
has increased significantly, as it
provides considerable value over
traditional datacentres – achieving
greater scalability, cost efficiency and
improved performance.1
Many organisations are now looking to quickly take
advantage of this value by migrating their existing
applications and workloads. But cloud migration
requires careful planning and strategy. One of the keys
to cloud success is to determine the optimal platform
and priorities for running business applications. Once
considered optional, these applications are now central to
infrastructure decisions and achieving company goals.
To enable successful migration, it’s important to
have a strong plan in place that covers the end-cloud
environment, training and, most importantly, the
readiness of your workloads and applications. To do this,
you need to determine how to successfully create the
initial technical plans and business justification, ensure
your workloads will run as expected and perform the
migration with limited impact on the business.
If you’re an IT manager running on-premises applications
and servers, this guide is designed to help you start their
migration to the cloud.
4
In this guide, you’ll discover:
Preliminary steps to consider when looking
at migration
Various approaches for rehosting, refactoring
and revising your workloads for the cloud
Tools you can use to accelerate your migration
project
Migrating to the cloud does not have to be difficult. With
the right tools and processes, your migration project
can be fast and friction free. Learn how methods such as
lift and shift rapidly move you to the cloud – reducing
cost immediately and enabling you to focus on future
cloud modernisation. Dive into assess, migration and
optimisation methods to help along the way, and start to
evaluate next steps in your continuing adoption of cloud
resources.
1
 y migrating to the cloud, you can save more than 80% over the
B
total cost of running a traditional solution on-premises. (Migrating
VMware to Microsoft Azure: Total cost of ownership guidance,
November 2017. Retrieved from: https://azure.microsoft.com/
mediahandler/files/resourcefiles/vmware-to-azure-migration-tco-
guidance/VMware_to_Azure_migration_TCO_guidance.pdf)
"One of our big objectives was to eliminate $3 million in capital
costs over about three years, and to reduce our operating costs by
approximately the same amount. At the same time, we wanted to
improve our quality of service. With Azure, we’re confident that we’ll
meet those goals.”

Jim Slattery, Chief Financial Officer, Capstone Mining2

Why migrate now?

At first glance, migration may seem like a technical Renewal of datacentre or hardware leasing. If you’re
decision, but at its core this is a business decision. currently extending your budget on renewing hardware
Ultimately, it begs a fundamental question: what’s driving or paying for datacentre locations for hosting, this is the
your business to migrate to the cloud, and why now? perfect time to look toward cloud migration. The cloud
can make these necessities a thing of the past by enabling
The benefits of cloud are universal – reduced running
a cloud vendor to host these services for you.
costs, faster modernisation capabilities and increased
security. But often, specific initiators first kick off the Renewal of licensing. Nearly everyone has an annual
discussion within an organisation, including: licensing agreement with their major IT providers. These
too require ample budget to ensure your virtualisation
Operational efficiencies and reduction of operating
and OS platform are sufficiently covered. The cloud can
expenses. Due to the reduction in hardware support,
help you here, providing a pay-as-you-go offering to
increased manageability and efficiency of process, you
reduce this cost.
can save an average of 20–30% on virtual-machine (VM)
resource configuration alone.³ Application development and modernisation. If
you’re in the software business, your resources are
Decreased time to market/release. By reducing
probably thinly spread. And using on-premises platforms
management overhead and freeing up budget, you
is likely not enabling you to adopt modern services.
can focus more time and effort on rapid software
The cloud provides an integrated platform for modern
and solution development. Faster deployment of
development, where development teams can increase
infrastructure as a service (IaaS) and platform as a service
speed by up to 33%.⁴
(PaaS) will enable your business to release faster and
more often. Ultimately, by migrating your current environment to
the cloud, you’re putting yourself in a better position
Support for scalability requirements that are more
to accelerate your business. By reducing costs and
cost effective. When you have to plan for peak usage
making management more efficient, a cloud platform
through on-premises systems, most of the time you’re left
can immediately impact your IT group’s ability to invest
with servers that are running at less than 20% utilisation.
back into core strategic projects, increasing security and
The cloud releases organisations from this model,
reliability while advancing application development.
enabling a scale-when-you-need-it approach.

Read the customer story:

2 

https://customers.microsoft.com/story/capstone-mining
3
VM sizing calculations are based on the Microsoft Azure
TCO Calculator, which is available at: https://www.tco.microsoft.com/
4
“The Shift to SaaS: A high-value opportunity for ISVs.” Keystone
Strategy. June 2017. Retrieved from: http://info.microsoft.com/rs/157-
GQE-382/images/EN-GB-CNTNT-Whitepaper-Keystone_Shift_to_SaaS_
5 June2017-en-gb.pdf
 02

How do you approach

cloud migration?

Sometimes cloud migration can
be simple, with only a few decision
points. However, your case may be
more complex, depending on how
many servers and virtual machines
you use.
Your migration could require you to run parallel and
iterative migration processes as you progressively move
your applications and workloads to the cloud.
Whether your migration is simple or complex, it’s helpful
to think of the basic elements of the process. Migration
can be boiled down to three main phases, as illustrated in
the model below.

PaaS SaaS
Lift and Cloud
shift optimisation

Existing App
on-premises modernisation
apps
IaaS

6
 PaaS
SaaS
Rearchitect
Rebuild

Refactor

IaaS Re-host Replace

Rehost. Also referred to as ‘lift and shift’, this stage

App data
entails migrating your physical servers and virtual
machines as-is to the cloud. By simply shifting your
Code
current server environment straight to IaaS, you reap the
benefits of cost savings, security and increased reliability.
Remains App configuration Managed by
In the new rehosted cloud model, hardware and same customer
operating systems you previously managed yourself are
App platform
now managed by the cloud provider. All other aspects of
the workload or application remain the same. This is the
Containers
most popular migration approach, as it lets organisations
move quickly, with little risk or impact and receive
OS
immediate benefits. It also allows customers to see lower Managed by
cloud provider
total cost of ownership (TCO) faster, enabling investment
Hardware
back into the migration process to evolve through the
model.
Figure A5

App data

‘Five Options for Migrating Applications to the Cloud: Rehost, Refactor, Revise, Rebuild or Replace.’ Jeff Woods. Gartner:
5

The Future of IT Conference. October 2011. https://gartnerinfo.com/futureofit2011/MEX38L_A2%20mex38l_a2.pdf

Code Managed by
customer
New/Updated
App configuration
7
App platform
 App data

Refactor. Also known as ‘repackage’, this stage involves

App data
Code
using additional cloud provider services to optimise
the cost, reliability and performance by refactoring
Remains Code
App configuration Managed
Managed by
by
your applications. In lift and shift, you were only taking
same customer
customer
Updated
advantage of the provider-managed hardware and
App configuration
App platform
OS, but in this model you also take advantage of cloud
services to drive down cost.
App platform
Containers
You continue to utilise your current application as-is, with
some minor application code or configuration changes,
Containers
OS
and connect your application to new infrastructure Managed
Managed byby
cloud
cloud provider
provider
services such as containers, databases and identity
OS
Hardware
management systems. By employing modernised services
in this scenario, you can lower cost and management.
Hardware

Figure B5

Rearchitect. This is also known as ‘redesigning’ an

App data
application to modernise it – that is, to transform it with
a modular architecture. Rearchitecting is modifying or
Code Managed by
extending an existing application’s code base to optimise customer
New/Updated
it for a cloud platform and for better scalability.
App configuration
Cloud provider services can be used directly as backend
services for modern apps, which are highly scalable and App platform
reliable. This is likely the most time-consuming way to
migrate an app to the cloud because it requires app Containers
code changes. One example of rearchitecting would be Managed by
cloud provider
decomposing a monolithic application into microservices OS
that work together and readily scale on Azure. Another
example would be rearchitecting a SQL Server database Hardware
to make it a fully managed Azure SQL Database.

Figure C5

App data

Code Managed by
customer
Updated

App configuration

‘Five Options for Migrating Applications to the Cloud: Rehost, Refactor, Revise, Rebuild or Replace.’ Jeff Woods. Gartner:
5

App platform
The Future of IT Conference. October 2011. https://gartnerinfo.com/futureofit2011/MEX38L_A2%20mex38l_a2.pdf

Containers
8 Managed by
cloud provider
OS
Rebuild. Revise the existing application by aggressively
App data
adopting PaaS or even software as a service (SaaS)
services and architecture. The process encompasses major
Code
revisions to add new functionality or to rearchitect the
application for the cloud.
Remains App configuration Managed by
An example of this stage would be code redesign to same customer

decompose the original application into smaller chunks, App platform

and then deployment using modern cloud provider
services. Containers

OS
Managed by
cloud provider
Hardware

Replace. This refers to moving or discarding an existing

Transform App data
application and replacing it with commercial software
delivered as a service, or SaaS. SaaS provides a complete
App data
Code
software solution that you purchase on a pay-as-you-go
basis from a cloud service provider.
Code Managed by
App configuration
customer
When you choose this option, all underlying New/Updated
infrastructure, middleware, app software, and app App configuration
App platform
Managed by
cloud provider
data are managed by service providers and located in
their datacentres. The service provider manages the App platform
Containers
hardware and software and, with the appropriate service
agreement, will ensure the availability and security of the Containers
OS
app – as well as your data. SaaS allows your organisation Managed by
cloud provider
to get quickly up and running with an app at minimal OS
Hardware
upfront cost. Typically, you migrate existing data to the
SaaS environment. Application data import/export is Hardware
achieved with an API or configuration/admin console.

App data

Code Managed by
customer
Updated

App configuration

App platform

9
Containers
Managed by
cloud provider
Lift and shift is the most common (and easiest) first step,
enabling you to move quickly to the cloud. Through
process discovery efforts, you can easily map the next
best steps for each workload based on goals, effort
and complexity. An advantage of this approach is that
it enables you to sustain parallel migration efforts.
And, as your IaaS projects continue, you can easily start
modernising certain applications to PaaS and even
SaaS options.
While the migration evolution model shows a potential
step-by-step journey for moving workloads from on-
premises to the cloud, the model also suggests that
workloads could start at different pivot points. Where you
begin depends upon the complexity of the workload and,
ultimately, what you want out of it. For example, if you
have a simple web application hosted in Windows Server
Internet Information Server (IIS), connected to a database,
then it may make sense to proceed directly to the cloud
optimisation stage. In this case, you could migrate
that application using Azure App Service and Azure
SQL Database. To do this, you’d need to have a deep
understanding of the application, including its complexity
and dependencies.
No matter which option you choose, you need a solution
that provides a smooth and easy cloud adoption, so
that you can migrate at your own pace. This requires
a cloud provider (and core partners) who can deliver
a comprehensive set of tools, methods and offers for
helping with migration and reducing risk. Most of all,
this solution should offer a simple process that’s easy
to follow.

With these goals in mind, Microsoft recommends a simple three-step

migration process for moving to the cloud:

Assess. Identify and inventory your
on-premises resources, such as
applications and workloads, to plan
where your Azure migration should
start.
Migrate. Migrate smarter and faster Optimise. Fine-tune your resources
with flexible, powerful tools, while to strengthen security, improve
ensuring minimal business impact. performance and maximise return
on investment (ROI).

10
With this straightforward framework, you get a proven
approach to migration. This process provides a clear
picture of your entire application and workload
portfolio, the best way to configure your applications
and workloads to achieve migration, convenient tools to
ensure low-impact transfer and ongoing performance
and cost optimisation.
But before migration begins, how do you ensure that
your virtual datacentre in the cloud is ready to receive
Before you migrate
Prior to migration, you essentially
need to build a virtual datacentre in
your cloud, including connectivity,
networking, storage and identity –
elements comparable to your
on-premises environment.
By building this virtual datacentre, you can ensure that
your applications will continue to support the business
after migration, without surprises.
Building this in the cloud is a streamlined process, as
you can leverage components available there for fast
assembly and configuration. Many excellent tools as
well as guidance can help accelerate this step, but below
are some of the basics – and why they’re important to
migration.
11
your workloads? As you’re planning to run your most
critical applications there, you want to be confident
that your cloud foundation is solid. If you already have
workloads or DevTest running on the cloud, then you
may have an environment and connectivity ready to go.
However, if you’re new to the cloud or only have a basic
set-up there, then you’ll need to perform a few steps
before migration.
Identity
As with your on-premises environment, you need a way
for users to be identified and authenticated to ensure
secure access to your new cloud resources. In most cases,
this means running Azure Active Directory or a similar
solution. You could simply extend your on-premises
identity to the cloud to support the migrated workloads.
However, many cloud providers have built-in identity
solutions provided as a service. These can also integrate
with your on-premises identity systems to enable single
sign-on for user convenience. As your workloads will
now be stored within the cloud, you’ll also want seamless
access for users between your existing on-premises
systems and the systems you’ve migrated.
To ensure a harmonious user experience – while
supporting your expected growth into more advanced
cloud services – a good investment would be a cloud-
based identity management solution connected to your
on-premises environment.
Storage
Another necessity in your cloud is a storage platform
that will meet the expected performance needs of your
migrated workloads. On-premises, this platform is usually
through network attached storage (NAS) or storage area
network (SAN) systems. In the cloud, virtual storage is
often through blobs or page blobs, depending on the
type of data being stored.
Fortunately, you can choose from many storage types to
guarantee reliable performance, including performance
and access levels, backup, geographical replication
and disaster recovery. Working with a virtual storage
platform means you also have a finer degree of control
and configuration. You can easily configure exact storage
requirements for each workload as needed, without
having to worry about physical architecture. Azure
even offers managed storage, which takes the pain out
of storage configuration. It requires you to enter just a
few details before acquiring up to thousands of secure,
reliable and available disks for your migration project.
Networking
Networks are crucial. They are the figurative backbone
of the datacentre. Moving to the cloud requires a new
way of networking, since you’re no longer reliant on
physical topology. In fact, you can now overcome
physical boundaries with a single subnet, which
simplifies networking communication.
When moving applications from on-premises to
the cloud, you may want to keep them in the same
networking subnets and even IP address ranges to
ensure a seamless migration. Virtual networking can
support this and merge with your on-premises physical
networking architecture as needed. This ensures that your
applications can continue to use the network topology
they were built upon, further easing migration.
12
Connectivity
During migration, you will move vast amounts of data.
However, you’ll still be moving data even after the bulk
of your critical workloads are hosted in the cloud. As
such, you should consider a more dedicated connectivity
option to help with the data transfer and ultimate end-
user experience.
You may currently have virtual networks set up, possibly
using Internet or site-to-site VPNs to connect to your
cloud environment. While this works well for smaller
deployments, you’ll need a new approach to run an
entire enterprise both during and after migration.
Azure ExpressRoute, for example, is a compelling tool to
use for this. It ensures both performance and security –
especially in the initial heavy lifting of virtual machines
to Azure, which involves considerable data flowing
across the wires. ExpressRoute enables a faster, private
connection to Azure.
Azure Data Box is another option for migrating large
amounts of data – when you’re limited by time, network
availability or cost. Data Box is a physical device that you
connect to your network. You load your data to it by
using standard NAS protocols. Then you simply send it to
Microsoft through a delivery transport service. Your data
will then be securely loaded into your Azure environment.
Readiness
While many core management skills can migrate to the
cloud, there are some key differences. You’ll need to get
up to speed on new skills. However, training takes time
away from performing core duties, can be costly and,
frequently, doesn’t properly build on what you already
know.
Azure Essentials can help solve these problems by
means of unique learning paths focused on job roles.
This readiness tool offers simple online training in bite-
sized pieces, practical labs and assessments to test your
knowledge. It’s the fastest way for your team to grow
their skills and, best of all, it’s free.
Beginning your migration
Now it’s time to begin your migration journey to the
cloud. This eBook focuses on a migration process where
you would be moving the bulk of your applications
and workloads running in virtual machines to IaaS (and
plan to modernise after that has been performed). The
environment in this scenario is set up in Azure with Azure
Active Directory (linked to on-premises AD) for identity
management, with managed disks ready to receive the
data and virtual networks deployed.

13
 03

Assess

In migrating to the cloud, you first

need to get a better understanding of
your applications, how many servers
and/or virtual machines you have
and how you’ll need to plan to move
them to the cloud. “We don’t want to be in the
Uncertainties about the total savings and perceived datacentre business; we’re in
complexity can get in the way of taking this step. Many
the thread business. We plan
organisations have come to realise that moving existing
workloads to Azure can yield significant benefits. to move 90% of our global
Justifying the investment requires confidence that you’ll datacentre infrastructure into
save a significant amount on operational costs and that
your current workloads will work as expected in the cloud.
Azure, and we’re at about
Many workloads can run immediately on Azure without 75% now. The only things we’ll
modification, while other workloads that have operational leave on-site are a few domain
and application dependencies in an on-premises
environment require further analysis and planning. If your
controllers and file/print
applications are made up of multiple servers or virtual servers.”
machines, then consolidated planning must be done
to identify these and shift them to the cloud. This is not
a manual process, and you’ll need intelligent planning Richard Cammish, Chief Information
tools to do it. Similarly, getting accurate cost comparisons Officer, Dillon Gage6
can be challenging when you’re estimating the load and
cloud VM series type. Without automated analysis to map
on-premises capacity to VM instance, your estimations
may fall short – causing performance issues. Or your
Read the customer story: https://customers.microsoft.com/en-us/story-
6

estimations may go too far – stretching your budget. precious-metals-dealer-ups-it-reliability-trims-costs

14
Assess in four steps
Technical and business planning for migration comes down to four straightforward steps:
Assessment of on-premises Identification of
applications and servers application and server
dependencies
Assessment of on-premises applications
and servers
It’s likely that your organisation runs hundreds – if not
thousands – of servers and virtual machines. While your
current management tools may represent these clearly,
to kick-start any migration you’ll need an assessment
mechanism that can feed data into subsequent steps.
Discovering servers and virtual machines is usually a
straightforward process. It relies on interaction directly
with the endpoint (using an agent) or managing
hypervisor (such as vSphere or Hyper-V).
Ultimately, the goal of the assess phase is to collect
server and application information, including type,
configuration, usage and applications that may be
running.
Configuration analysis
The assess process enables you to ensure that each
workload will function on your cloud platform. Through
the collected analysis, assessment tools will be able to
provide metrics on the compatibility of the workload in
the cloud. For example, is the workload OS supported?
Or are there specific hardware dependencies that may
not be replicated in a cloud environment (such as running
an UEFI boot, which is larger than a 4 TB data disk size)?
Configuration analysis should tell you which workloads
15
Configuration Cost planning
analysis
Identification of application and server
dependencies
Once discovery is complete, you’ll need to map any
dependencies or communication between your servers
(and applications). This is critical because when migrating
an application, you need to know all the servers and
processes the app is using.
Many tools provide server dependency mapping, but
don’t extend to application dependencies. To ensure a
full picture of all communication between workloads, you
need a tool that will do both. This will allow you to create
visual maps of all your applications and workloads, which
enables their interaction as a single entity for costing,
configuration analysis and eventually migration.
will migrate with no modifications, which workloads
might require basic modifications to comply and which
workloads are not compatible in their current formation,
as well as provide guidelines to remediate potential issues
or recommend configuration changes.
Cost planning
The final step of the assess phase is collecting resource
usage reporting (such as CPU, memory and storage).
This is important, as on-premises virtual machines are
often over-provisioned, but actually utilised under 20%.
If you were to take the physical configuration of your
on-premises server and map this to an IaaS VM series
type, you would likely find that you were paying for
more performance and scale than you need.
For further information
on these tools, explore
the following resources:
Azure Migrate. Take advantage of discovery,
assessment, guidance, insight and mechanisms
for cloud migration through this free integrated

Since the cloud is costed as a usage model, you should Azure service.

ensure your choice meets both performance and › Learn more about Azure Migrate

economic targets. The goal in any cloud model is to

drive your virtual machines to at least 90% utilisation,
Azure migration partners. Accelerate your
while making sure you meet performance and reliability
migration with experienced assessment partners,
goals. Through historic resource analysis, intelligent cost
who provide many options for your unique
analysis tools can determine the actual usage of your
environments.
workload and suggest the best cloud IaaS VM series
› Learn more about Azure migration
to use.
technology partners

Tools for assessment

Azure Hybrid Benefit. Use your on-premises
Many tools in the Azure ecosystem enable you to Windows Server licences with Software Assurance
tackle these needs simultaneously. As part of the Azure to make big savings on Azure. With this benefit,
subscription for all customers, Microsoft offers Azure for each licence Microsoft will cover the cost of
Migrate to provide automation for the assess phase. the OS (on up to two virtual machines!), while you
Azure Migrate is a great fit for customers who have pay only for base compute costs.
virtualised servers in VMware. Customers with servers in › Learn more about Azure Hybrid Benefit
other environments should leverage assessment tools
from the Azure migration partner ecosystem. These tools
Azure for Windows Server apps.
can also help you with usage characteristics like CPU,
Save up to 82% on the pay-as-you-go price of
memory and storage to equivalent Azure environments,
running Windows Server apps in Azure Virtual
giving you the technical and business reporting needed
Machines when you combine the Azure Hybrid
to continue your migration plans. Using these tools will
Benefit for Windows Server with the upcoming
help you maximise the benefits of moving to Azure, as
Reserved Virtual Machine Instance offering.
well as identify where programs such as Azure Hybrid
› Learn more about running Windows Server
Benefit best fit into your migration to save further
apps on Azure
budget. With Azure Hybrid Benefit, you can utilise your
on-premises Windows Server licences with Software
Assurance when migrating and save up to 40% in Azure
VM runtime costs.7

7
 zure Hybrid Benefit. Retrieved from:
A
https://azure.microsoft.com/pricing/hybrid-benefit/

16
 04

Migrate

Once you have completed your
discovery and assessment, it’s time
to prepare for the next step: cloud
migration.
This is where, after you have landed on your migration
goals and gathered all requirements and constraints, you
can choose the best method of migration.
Earlier in this eBook, we overviewed the migration
evolution model that showed workloads progressing
through phases of lift and shift, cloud optimisation and
eventually app modernisation.
During the migration effort, you’ll determine the
approach that best meets your requirements. This is
best addressed per application. Essentially, in this phase,
you’re physically moving your workloads and applications
(including their data) to the cloud and planning to retire
the on-premises versions. Every organisation will have
a different approach and mixture of using rehosting,
refactoring, rearchitecting or rebuilding for their
applications.
This eBook focuses on the rehost approach, moving
applications running on traditional servers and virtual
machines to Azure IaaS. In many cases, organisations will
start with lift and shift to drive rapid migration and early
cost savings. Lift and shift involves no change in your app
or workload framework or architecture; it simply means
exchanging hardware and OS management with the
cloud environment. This approach requires confidence
regarding two key issues. Can your workload be easily
migrated, without too many manual steps? Will your
workload function as expected in the cloud? As such,
several decision points come into play based on what’s
being moved, and especially how (or if) you want to
access it while the migration is taking place.
The lift and shift method most often employed for server
or VM migration is real-time replication, because of its
flexibility and capability in staged migration. Real-time
replication allows the workload to remain online and
accessible during the migration. And, as you’ll see in the
next section, modern tools enable the system to cleanly
migrate real-time data even when the system is actively
being used.

PaaS SaaS PaaS

SaaS
Lift and Cloud Rearchitect
shift optimisation
Rebuild

Refactor

Existing App
on-premises modernisation
apps
IaaS IaaS Re-host Replace

17
Real-time replication
Real-time replication involves setting up a copy of
Cloud provider
the workload in the cloud and allowing asynchronous
replication to keep the copy and the workload in sync.
VM

This means that while you’re building and executing your VM

migration plans, any data or server updates are synced

VM

VM

between the copies. On-premises datacentre

This model also enables groups of virtual machines vSphere or Hyper-V Host Firewall

to be connected, as in a multi-tiered application or VM

Replication

workload. This is important when testing and the VM

final migration cutover begins. When the system is VM

aware of the connections and dependencies between VM

virtual machines, you can create plans to ensure the

VMs are bought up in the correct order when starting.
For example, with a simple web app, your database
source needs to be available before the application
runtime begins.
Using your assessment plans as a guide – and your
migration tool of choice – you can configure each VM
to replicate to the correct VM instance in your cloud
provider. This is also the point when you should define
the storage and network connections that you set up
initially during the environment’s creation. Most tools
have a mechanism to define the replication timeframe
(usually between 30 seconds and 15 minutes). This will
be based on your network capability and latency.
Many tools also support application-aware replication
automatically. Microsoft applications (such as SharePoint,
Dynamics, SQL Server and Active Directory) and apps
from other vendors (including Oracle, SAP, IBM and Red
Hat) can be migrated with application-aware replication,
which ensures the source data consistency before
replication. Initial replication is also bandwidth intensive,
and mechanisms discussed earlier – like ExpressRoute and
Data Box – can assist with this. It’s something to consider
when planning your migration timeline.

Testing
Testing is integral to ensuring system health before final Migration tools can also do the final launch in your cloud
cut-over. Many migration tools include options to start up and turn off the on-premises version. In some cases, you
your set of VMs in an isolated environment, which allows may have to update some DNS records for the new cloud-
you to mimic the production environment in the cloud. based workloads. However, if you migrated to use DNS
This means that you can fully test the application without in the cloud as part of your initial environment setup, this
affecting either the on-premises or cloud production may happen automatically.
versions. Once replication is complete, simply start your
application or workloads using the isolated environment
option, while taking time to test your start-up script or
runbook for any errors. When you’re fully satisfied that
both function as expected, it’s time to perform the final
cut-over.

18
“To expand globally, we can simply clone the infrastructure that
we have running in our US Microsoft Azure datacentre to Azure
datacentres in Asia and elsewhere.”

Tom Grounds, Chief Information Officer, Coats8

Taking further advantage of your cloud provider to drive application innovation

As you migrate your existing virtual machines in Azure,

this is also the perfect time to continue on the path to
application modernisation. To do this, you can take
$$$ On-premises
advantage of your cloud provider, delivering even further
cost savings and flexibility.

As you may remember from the migration evolution

model, cloud optimisation is the next logical step $$
IaaS
following lift and shift, or ‘rehosting’, for your workloads,
PaaS
as you’ll find many of them can take advantage of PaaS
services.
15-20% 35%
savings over savings over
The PaaS services of immediate interest are containers, $ on-premises⁹ on-premises, 15%
app services and database services. Why look at cloud over VMs⁹

optimisation so soon after migrating? The answer is

simple. Now that you have done all the hard work in *Based on assumptions using the Azure Cost Calculator

assessing, analysing and migrating to the cloud, you’ve

made it easier to take the step to PaaS. Plus, you can get
more cost benefits through reduced management and
operating expense reduction.
A clear place to start is containerisation. Containerisation
provides an OS-level virtualisation method
used to deploy and run distributed applications
without launching an entire VM for each app. With
containerisation, you get immediate savings on
operational expenditure due to a reduced footprint in
the cloud. Remember that the smaller your footprint, the
more economical it is. Containerisation rapidly moves
compatible apps to containers (out of large VMs) with
no code changes. It immediately provides you with the
benefit of running multiple apps with no affect on the
end user experience.
The next consideration in optimisation is moving to
specific PaaS services. There are a lot of options to choose
from, but two to start with are app services and database
services because of the ease of the migration activity. In
many cases, web and mobile applications can migrate
to App Service with little refactoring effort.
By moving to PaaS for your suitable apps and databases,
you’re significantly lowering costs by reducing your
footprint and management requirements. You can save
an additional 15–20% or more by migrating workloads
and applications to PaaS, over and above the savings you
are making today with IaaS.
It’s important to note that there will be an initial
investment of time, effort and budget to move to more
advanced PaaS services. In some cases, you’ll have
8
Read the customer story:
https://customers.microsoft.com/story/coats
9
Based on the Microsoft Azure Cost Calculator. Retrieved from:
https://www.tco.microsoft.com/

19
a simple migration where the application is ‘cloud ready’,
but in others configuration changes and code updates
may be required. Fortunately, there are tools available
For further information
that enable you to analyse the code and determine on these tools, explore
the effort required to move to App Service.
the following resources:
For your databases, this is a straightforward process.
Azure provides options for database PaaS services,
Azure Site Recovery. Migrate physical servers
including Azure SQL Database, where you are hosting
or VMs from on-premises to Azure, or support
the data on a full database as a service (DBaaS) platform.
custom replication timing, isolated testing and
Azure SQL Database enables you to host database data
final migration cut-over. You can replicate and
in a service, reducing your database management costs,
migrate your VMs to Azure using right-sizing
but has some limitations compared to a full SQL Server
recommendations for migration as part of the
deployment. If your applications need functionality that
assessment phase and take advantage of Azure
may not be available in Azure SQL Database, then SQL
Hybrid Benefit when the initial replication occurs
Database Managed Instance is recommended. Azure SQL
(as a configuration option). This tool saves you
Database Managed Instance provides the full platform
time, as you don’t need to retroactively go back
experience of SQL Server, but with the underlying OS
and perform these tasks for each VM.
and SQL service fully managed by Azure.
› Learn more about Azure Site Recovery

Tools for migration Azure Database Migration Service. Migrate

As the migration of servers and virtual machines is your existing application database to Azure
different for everyone, multiple tools are available to as a VM, Azure Managed Instance or directly
support your needs. These range from the Microsoft- to Azure SQL Database.
provided Azure Site Recovery (ASR) to third-party tools. › Learn more about the Azure Database
Third-party tools are valuable alternatives when you Migration Service
have specific needs not covered by ASR. For example,
while there are some OS types ASR can’t migrate,
Azure migration partners. Accelerate your
various partner tools can support these efforts.
migration to Azure with partners offering
Database migration is uniquely supported by the Azure technology solutions and the expertise to help.
Database Migration Service (DMS). By using the DMS › Learn more about Azure migration
migration workflow, you will be able to move your on- technology partners
premises databases to Microsoft Cloud. DMS enables
schema and data migrations from SQL Server to Azure,
including SQL Server on virtual machines and Azure SQL
Database.

Alternatively, you may have other specific needs, like

rapid migration (migrating over a hundred VMs per
day), where normal replication may not be sufficient.
In this case, tools can assist in migrating the run-time
to Azure first while leaving the storage on-premises.
Then, over time, the storage is replicated. Many options
can meet your unique migration needs. Learn more
at Azure Migration Centre.

20
 05
Optimise
Once you’ve implemented your
cloud migration strategy, you’ll want
to ensure that you’re successfully
taking advantage of the cloud’s
performance, scalability and cost-
saving benefits.
This will enable you to only pay for the services and
resources you use, achieve a greater ROI and receive
additional savings by taking advantage of the latest cloud
capabilities. This is also the best time to start looking at
new services for modernising your application, migrating
to PaaS and even SaaS, where applicable.
On-premises tools are not built for cloud scale and
agility. Plus, they’re simply not aligned with the
new usage models enabled by the cloud. Continual
optimisation is a critical third step in your migration
journey. Optimisation targets two main areas – ensuring
peak performance and continual cost efficiency.
Once you migrate, you’ll also want to make sure to keep
your virtual machine continuously secure, protect your
data and monitor your cloud health. And that’s easy to
do with Azure – once you get an understanding of the
full suite of controls and capabilities available to you.
21
Secure cloud resources
Ensuring strong security for your cloud-based resources
is a shared responsibility between you and your cloud
provider. Azure is built on a foundation of trust and
security, compliance, privacy and transparency. The
Azure platform provides a secure foundation to host
your infrastructure with built-in security controls and
capabilities to help you further protect your data and
applications.
Azure Security Centre provides unified security
management and advanced threat protection across
hybrid cloud workloads. The Security Centre enables
you to take advantage of several capabilities, including:
› Centralised policy management. Ensure compliance
with company or regulatory security requirements by
centrally managing security policies across all your
hybrid cloud workloads.
› Continuous security assessment. Monitor the
security of machines, networks, storage and data
services and applications to discover potential security
issues.
› Actionable recommendations. Remediate security
vulnerabilities before they can be exploited by
attackers with prioritised and actionable security
recommendations.
› Advanced cloud defences. Reduce threats with just-
in-time access to management ports and whitelisting
to control applications running on your VMs.
› P
 rioritised alerts and incidents. Focus on the most
critical threats first with prioritised security alerts and
incidents.
› Integrated security solutions. Collect, search
and analyse security data from a variety of sources,
including connected partner solutions.
Protecting data
Monitoring cloud health
As with any system, monitoring is important to drive both
proactive and reactive analysis. Azure provides many
monitoring services targeted at applications, workloads
and core service health to ensure you have full visibility
into your current system status – plus, access to important
data when working with a break-fix situation. In Azure,
you can use either basic or premium monitoring services.

Basic monitoring provides fundamental, required

Azure ensures workloads and data are fully backed up
monitoring across Azure resources. These services need
and protected from disasters, while providing encryption
minimal configuration and collect core telemetry that the
of stored data for internal and customer security. Azure
premium monitoring services use.
can also automatically encrypt your stored data – while
allowing full accessibility to all applications and users. › A zure Monitor. Azure Monitor enables basic
monitoring for Azure services by allowing the
› V
 irtual machine disk encryption. Azure Disk
collection of metrics, activity logs and diagnostic
Encryption enables encryption of Windows and Linux
logs. For example, the activity log tells you when
Azure Virtual Machine disks. Azure Disk Encryption
new resources are created or modified. Metrics are
uses the industry standard BitLocker feature of
available that provide performance statistics for
Windows and the dm-crypt feature of Linux to
different resources, and even the operating system,
provide volume encryption for the OS and the data
inside a virtual machine. You can view this data with
disks. The solution is integrated with Azure Key Vault
one of the explorers in the Azure portal, send it to
to help you control and manage the disk encryption
Azure Log Analytics for trending and detailed analysis
keys and secrets in your key vault subscription, while
or create alert rules to proactively get notifications of
ensuring that all data on the virtual machine disks is
critical issues.
encrypted at rest in your Azure storage.
› Service Health. It’s important to be aware of any
› Virtual machine backup. Azure Backup is a scalable
issues with Azure services concerning any dependent
solution that protects your application data with
applications. Azure Service Health identifies issues
zero capital investment and minimal operating costs.
with Azure services that might affect your application.
Application errors can corrupt your data, and human
Service Health also helps you plan for scheduled
errors can introduce bugs into your applications.
maintenance.
With Azure Backup, your virtual machines running
Windows and Linux are protected. › A zure Advisor. Azure Advisor constantly monitors
your resource configuration and usage telemetry. It
› A zure Site Recovery. An important part of your
then gives you personalised recommendations based
organisation’s business continuity and disaster
on best practices. Following these recommendations
recovery (BCDR) strategy is figuring out how to keep
helps you improve the performance, security and
corporate workloads and apps up and running when
availability of the resources that support your
planned and unplanned outages occur. Azure Site
applications.
Recovery helps orchestrate replication, failover and
recovery of workloads and apps so that they are
available from a secondary location if your primary
location goes down.

22
Premium monitoring services build on basic
monitoring and provide powerful analytics with collected
data to give you unique insights into your applications
and infrastructure. Plus, they present you with data in
the context of scenarios that are targeted to different
audiences.
Many premium management solutions are packaged
sets of logic that provide insights for an application or
service. They rely on log analytics to store and analyse the
monitoring data that they collect. Azure Log Analytics
enables deeper visibility into your hybrid IT environment
and allows you to diagnose performance issues from an

› A
 pplication Insights enables you to monitor advanced analytics portal in one click.

the availability, performance and usage of your Azure Log Analytics enables you to:
application, whether it’s hosted in the cloud or
› Analyse data. In Log Analytics, you can leverage
on-premises. By instrumenting your migrated or
log searches by constructing queries to analyse the
rearchitected application to work with Application
collected data, using pre-existing dashboards that
Insights, you can quickly identify and diagnose errors
you can customise with graphical views of your most
without waiting for a user to report them. With the
valuable searches. Once you have a defined collection
information that you collect, you can make informed
of operational data from your Azure virtual machines
choices on your application’s maintenance and
and activity logs, you can perform powerful searches.
improvements. Application Insights stores its data in
› Visualise data. Log Analytics dashboards can visualise
a common repository where it can take advantage of
all your saved log searches, giving you the ability to
shared functionality such as alerts, dashboards and
find, correlate and share IT operational data in your
deep analysis with the Log Analytics query language.
organisation.
› Service Map provides insight into your IaaS
› Alert data. Alerts in Log Analytics identify important
environment by analyzing virtual machines with
information in your repository. They are created
their different processes and dependencies on other
by alert rules that automatically run log searches
computers and external processes. It integrates
at regular intervals and match certain criteria. With
events, performance data and management solutions
Alert Action, you can perform advanced actions with
in Log Analytics. You can then view this data in the
alerts, such as create an email notification, launch an
context of each computer and its relation to the rest
automation runbook or create an incident record in
of your environment.
your ITSM incident management system.
› Network Watcher provides scenario-based
monitoring and diagnostics for different network
scenarios in Azure. It stores data in Azure metrics
and diagnostics for further analysis.

23
Continual cost efficiency and optimisation Tools for optimisation
Performance monitoring can help you achieve beneficial
cost optimisation. In the initial assessment, you For further information on
performed right-sizing for your on-premises workload
optimisation tools, explore
based on a point in time. Once those workloads are
moved to Azure, their usage may change. For example, the following resources:
if you move a moderately used app from on-premises
to Azure, the initial assessment phase may recommend Security with Azure Security Centre. With
for the middle-tier VM a D2v3 instance (2 vCPU and Security Centre, you can apply security policies
8 GB RAM). However, after six months, the use of that across your workloads, limit your exposure to
application might decline, so you’d want to downsize threats and detect and respond to attacks.
to a lower VM instance to reduce costs. › Learn more about the Azure Security Centre
Azure Cost Management shows you usage and costs so
that you can track trends, detect inefficiencies and create
Azure Cost Management. Get detailed visibility
alerts. All usage and cost data are displayed in intuitive
into the resource costs of running systems
dashboards and reports. With built-in cost management,
on Azure through this service that is tightly
you can continually monitor for CPU and memory usage,
integrated with Azure operations.
enabling recommendations for virtual-machine instances
› Learn more about Azure Cost Management
that can be further right-sized. These services can also
help you monitor for over-utilised VMs and up-size as
needed to ensure performance service level agreements Azure Log Analytics. This service centralises log
(SLAs). Plus, these services can help you discover data from multiple systems in a single data store.
underutilised VMs for potential downsizing. For example, › Learn more about Azure Log Analytics
Azure cost optimisation can provide a regular view of
your current virtual-machine total utilisation.
Monitoring in Azure. Get detailed, up-to-
With a quick glance, you can determine the number of
date performance and utilisation data, access
virtual machines that are consistently under-utilised
to the activity log that tracks every API call and
(that is, running below 90%). Then, with the Azure
diagnostic logs that help you debug issues in
cost optimisation sizing opportunities, you can find
your Azure resources. Management solutions are
recommendations for which VMs should be actioned,
available from Microsoft and partners to provide
as well as the suggested instance change (including
monitoring for various Azure and third-party
potential annual savings).
services.
As you continue to utilise your new IaaS environment, › Learn more about Azure Monitoring
targeting maximum cost savings through Azure Reserved
VM Instances (RI) becomes attractive. Reporting
available in Azure Cost Management can recommend Azure migration partners. Accelerate your
the workloads that would benefit from RIs, maximising migration with experienced assessment partners,
your TCO. who provide many options for your unique
environments.
› Learn more about Azure migration
technology partners

24
 06
Summary
No matter why your organisation is
looking to move to the cloud, you
shouldn’t be put off by its seeming
complexity.
As you’ve learned in this eBook, breaking a migration
down into elemental steps will contribute to a more
successful migration. You can start this process
with a few applications and then expand to further
applications in your environment.
First, you need a clear plan that takes into careful
consideration your servers, VMs and workloads – and
what’s required for these to function in the cloud. Along
with this assessment, determine your true resource usage
and analyse any configuration dependencies for your
workloads. Then, when you go forward with migration,
ensure that your workloads in the cloud are in sync with
your on-premises system in real time. Along with this, you
want to test the health of your system, so that your final
cutover is smooth. Lastly, continue your cloud journey by
ensuring peak performance and cost efficiency through
cloud optimisation.
To learn more about options
to meet your unique migration
needs, visit the
Azure Migration Center
25
The benefits of migration will be immediately apparent
in your time and budget savings. The cloud will allow
you to be more agile and, in many cases, respond to
business needs faster. The cloud may even lower your
TCO by as much as 84%, freeing you to take that massive
savings and invest it back into your business to drive
modernisation faster. Plus, you can explore PaaS and SaaS
options, decreasing your TCO even more while expanding
your IT capability.
Whether you’re in the early stages of migration
assessment, or just starting to plan your approach,
keep in mind that migration can be easier with a
trusted provider like Microsoft. Through integrated
tools, a strong partner ecosystem, and rich guidance,
you can tread a well-forged path to minimise risk and
impact to your business.
“Migrating VMware to Microsoft Azure: Total cost of ownership
10
guidance.” Microsoft. November 2017. Retrieved from: https://
azure.microsoft.com/mediahandler/files/resourcefiles/vmware-to-
azure-migration-tco-guidance/VMware_to_Azure_migration_TCO_
guidance.pdf
Additional resources
Azure Migration Guide for
Windows Server
Choosing your database
migration path to Azure